Estimating intervention effectiveness: synthetic projection of field evaluation results.

In a 46-site, 5-year high-risk youth substance abuse prevention evaluation, effect sizes were adjusted using a meta-analytic regression technique to project potential effectiveness under more optimal research and implementation conditions. Adjusting effect size estimates to control for the impact of comparison group prevention exposure, service intensity, and coherent program implementation raised the mean effectiveness estimate from near zero (.02, SD = .21) to .24 (SD = .18). This finding suggests that adolescent prevention programs can have significant positive effects under optimal, yet obtainable conditions. EDITORS' STRATEGIC IMPLICATIONS: The authors present a meta-analytic technique that promises to be an important tool for understanding what works in multi-site community-based prevention settings. Researchers will find this to be a creative approach to model the "noise'' in implementation that may often overshadow the potential impact of prevention programs.